Training an automated circulating tumor cell classifier when the true classification is uncertain
Ontology highlight
ABSTRACT:
SUBMITTER:
PROVIDER: S-EPMC10873494 | biostudies-literature | 2024 Feb
REPOSITORIES: biostudies-literature
Similar Datasets
Project description:Verbal autopsy (VA) deals with post-mortem surveys about deaths, mostly in low and middle income countries, where the majority of deaths occur at home rather than a hospital, for retrospective assignment of causes of death (COD) and subsequently evidence-based health system strengthening. Automated algorithms for VA COD assignment have been developed and their performance has been assessed against physician and clinical diagnoses. Since the performance of automated classification methods remains low, we aimed to enhance the Naïve Bayes Classifier (NBC) algorithm to produce better ranked COD classifications on 26,766 deaths from four globally diverse VA datasets compared to some of the leading VA classification methods, namely Tariff, InterVA-4, InSilicoVA and NBC. We used a different strategy, by training multiple NBC algorithms using the one-against-all approach (OAA-NBC). To compare performance, we computed the cumulative cause-specific mortality fraction (CSMF) accuracies for population-level agreement from rank one to five COD classifications. To assess individual-level COD assignments, cumulative partially-chance corrected concordance (PCCC) and sensitivity was measured for up to five ranked classifications. Overall results show that OAA-NBC consistently assigns CODs that are the most alike physician and clinical COD assignments compared to some of the leading algorithms based on the cumulative CSMF accuracy, PCCC and sensitivity scores. The results demonstrate that our approach improves the performance of classification (sensitivity) by between 6% and 8% compared with other VA algorithms. Population-level agreements for OAA-NBC and NBC were found to be similar or higher than the other algorithms used in the experiments. Although OAA-NBC still requires improvement for individual-level COD assignment, the one-against-all approach improved its ability to assign CODs that more closely resemble physician or clinical COD classifications compared to some of the other leading VA classifiers.
| S-EPMC6480413 | biostudies-literature
Project description:Circulating tumor cells (CTCs) have the potential of becoming the gold standard marker for cancer diagnosis, prognosis and monitoring. However, current methods for its isolation and characterization suffer from equipment variability and human operator error that hinder its widespread use. Here we report the design and construction of a fully automated high-throughput fluorescence microscope that enables the imaging and classification of cancer cells that were labeled by immunostaining procedures. An excellent agreement between our machine vision-based approach and a state-of-the-art microscopy equipment was achieved. Our integral approach provides a path for operator-free and robust analysis of cancer cells as a standard clinical practice.
| S-EPMC6760523 | biostudies-literature
Project description:Circulating tumor cells, a component of the "liquid biopsy", hold great potential to transform the current landscape of cancer therapy. A key challenge to unlocking the clinical utility of CTCs lies in the ability to detect and isolate these rare cells using methods amenable to downstream characterization and other applications. In this review, we will provide an overview of current technologies used to detect and capture CTCs with brief insights into the workings of individual technologies. We focus on the strategies employed by different platforms and discuss the advantages of each. As our understanding of CTC biology matures, CTC technologies will need to evolve, and we discuss some of the present challenges facing the field in light of recent data encompassing epithelial-to-mesenchymal transition, tumor-initiating cells, and CTC clusters.
| S-EPMC5528969 | biostudies-other
Project description:BackgroundThe ability of circulating tumor cells (CTCs) to identify lung adenocarcinoma (LUAD) and lung squamous cell carcinoma (LUSC) could improve pathological diagnosis and the selection of treatments for non-small cell lung cancer (NSCLC). Previous studies have shown that deoxyribonucleic acid (DNA) methylation exhibits cell and tissue specificity. Thus, we aimed to explore the methylation status of CTCs in LUAD and LUSC and identify the potential biomarkers.MethodsWe first analyzed Infinium 450K methylation profiles obtained from The Cancer Genome Atlas and Gene Expression Omnibus. We then performed whole-genome sequencing of CTCs in tumor and matched normal lung tissues and white blood cells from 6 NSCLC patients.ResultsThe bioinformatics analysis revealed a NSCLC-specific DNA methylation marker panel, which could accurately distinguish between LUAD and LUSC with high diagnostic accuracy. The whole-genome sequencing of CTCs in NSCLC patients also showed 100% accuracy for distinguishing between LUAD and LUSC based on the CTC methylation profiles. To investigate the function of CTCs, we further analyzed similar and different methylation profiles between the CTCs and their primary tumors, and found very high similarities between the CTCs and their primary tumor tissues, indicating that these cells inherit information from primary tumors. However, the CTCs also displayed some characteristics that differed to those of primary tumor tissues, which suggest that CTCs acquire some unique characteristics after migrating from the primary tumor; these characteristics may partly explain the ability of tumor cells to evade immune surveillance.ConclusionsOur findings provide insights into the potential use of CTCs in the pathological classification of NSCLC patients. Our findings also show how CTC primary tumor inheritance and CTC evolution affect metastasis and immune escape.
| S-EPMC8902094 | biostudies-literature
Project description:While circulating tumor cell (CTC)-based detection of AR-V7 has been demonstrated to predict patient response to second-generation androgen receptor therapies, the rarity of AR-V7 expression in metastatic castrate-resistant prostate cancer (mCRPC) suggests that other drivers of resistance exist. We sought to use a multiplex gene expression platform to interrogate CTCs and identify potential markers of resistance to abiraterone and enzalutamide. 37 patients with mCRPC initiating treatment with enzalutamide (n = 16) or abiraterone (n = 21) were prospectively enrolled for CTC collection and gene expression analysis using a panel of 89 prostate cancer-related genes. Gene expression from CTCs was correlated with PSA response and radioclinical progression-free survival (PFS) using Kaplan-Meier and Cox regression analyses. Twenty patients (54%) had detectable CTCs. At a median follow-up of 11.3 months, increased expression of the following genes was significantly associated with shorter PSA PFS and radioclinical PFS: AR, AR-V7, PSA, PSCA, TSPAN8, NKX3.1, and WNT5B. Additionally, high SPINK1 expression was associated with increased PFS. A predictive model including all eight genes gave an area under the curve (AUC) of 0.84 for PSA PFS and 0.86 for radioclinical PFS. In comparison, the AR-V7 only model resulted in AUC values of 0.65 and 0.64.These data demonstrate that clinically relevant information regarding gene expression can be obtained from whole blood using a CTC-based approach. Multigene classifiers in this setting may allow for the development of noninvasive predictive biomarkers to guide clinical management.
| S-EPMC6612010 | biostudies-literature
Project description:PurposeDeep convolutional neural networks (CNN) are promising for automatic classification of dopamine transporter (DAT)-SPECT images. Reporting the certainty of CNN-based decisions is highly desired to flag cases that might be misclassified and, therefore, require particularly careful inspection by the user. The aim of the current study was to design and validate a CNN-based system for the identification of uncertain cases.MethodsA network ensemble (NE) combining five CNNs was trained for binary classification of [123I]FP-CIT DAT-SPECT images as "normal" or "neurodegeneration-typical reduction" with high accuracy (NE for classification, NEfC). An uncertainty detection module (UDM) was obtained by combining two additional NE, one trained for detection of "reduced" DAT-SPECT with high sensitivity, the other with high specificity. A case was considered "uncertain" if the "high sensitivity" NE and the "high specificity" NE disagreed. An internal "development" dataset of 1740 clinical DAT-SPECT images was used for training (n = 1250) and testing (n = 490). Two independent datasets with different image characteristics were used for testing only (n = 640, 645). Three established approaches for uncertainty detection were used for comparison (sigmoid, dropout, model averaging).ResultsIn the test data from the development dataset, the NEfC achieved 98.0% accuracy. 4.3% of all test cases were flagged as "uncertain" by the UDM: 2.5% of the correctly classified cases and 90% of the misclassified cases. NEfC accuracy among "certain" cases was 99.8%. The three comparison methods were less effective in labelling misclassified cases as "uncertain" (40-80%). These findings were confirmed in both additional test datasets.ConclusionThe UDM allows reliable identification of uncertain [123I]FP-CIT SPECT with high risk of misclassification. We recommend that automatic classification of [123I]FP-CIT SPECT images is combined with an UDM to improve clinical utility and acceptance. The proposed UDM method ("high sensitivity versus high specificity") might be useful also for DAT imaging with other ligands and for other binary classification tasks.
| S-EPMC10957699 | biostudies-literature
Project description:detailed information about circulating tumor cells (CTCs) as an indicator of therapy response and cancer metastasis is crucial not only for basic research but also for diagnostics and therapeutic approaches. Here, we showcase a newly developed IsoMAG IMS system with an optimized protocol for fully automated immunomagnetic enrichment of CTCs, also revealing rare CTC subpopulations. using different squamous cell carcinoma cell lines, we developed an isolation protocol exploiting highly efficient EpCAM-targeting magnetic beads for automated CTC enrichment by the IsoMAG IMS system. By FACS analysis, we analyzed white blood contamination usually preventing further downstream analysis of enriched cells. 1 µm magnetic beads with tosyl-activated hydrophobic surface properties were found to be optimal for automated CTC enrichment. More than 86.5% and 95% of spiked cancer cells were recovered from both cell culture media or human blood employing our developed protocol. In addition, contamination with white blood cells was minimized to about 1200 cells starting from 7.5 mL blood. Finally, we showed that the system is applicable for HNSCC patient samples and characterized isolated CTCs by immunostaining using a panel of tumor markers. Herein, we demonstrate that the IsoMAG system allows the detection and isolation of CTCs from HNSCC patient blood for disease monitoring in a fully-automated process with a significant leukocyte count reduction. Future developments seek to integrate the IsoMAG IMS system into an automated microfluidic-based isolation workflow to further facilitate single CTC detection also in clinical routine.
| S-EPMC8623084 | biostudies-literature
Project description:Introduction: Endothelial cells (ECs), being located at the interface between flowing blood and vessel wall, maintain cardiovascular homeostasis by virtue of their ability to integrate chemical and physical cues through a spatio-temporally coordinated increase in their intracellular Ca2+ concentration ([Ca2+]i). Endothelial heterogeneity suggests the existence of spatially distributed functional clusters of ECs that display different patterns of intracellular Ca2+ response to extracellular inputs. Characterizing the overall Ca2+ activity of the endothelial monolayer in situ requires the meticulous analysis of hundreds of ECs. This complex analysis consists in detecting and quantifying the true Ca2+ events associated to extracellular stimulation and classifying their intracellular Ca2+ profiles (ICPs). The injury assay technique allows exploring the Ca2+-dependent molecular mechanisms involved in angiogenesis and endothelial regeneration. However, there are true Ca2+ events of nearly undetectable magnitude that are almost comparable with inherent instrumental noise. Moreover, undesirable artifacts added to the signal by mechanical injury stimulation complicate the analysis of intracellular Ca2+ activity. In general, the study of ICPs lacks uniform criteria and reliable approaches for assessing these highly heterogeneous spatial and temporal events. Methods: Herein, we present an approach to classify ICPs that consists in three stages: 1) identification of Ca2+ candidate events through thresholding of a feature termed left-prominence; 2) identification of non-true events, known as artifacts; and 3) ICP classification based upon event temporal location. Results: The performance assessment of true-events identification showed competitive sensitivity = [0.9995, 0.9831], specificity = [0.9946, 0.7818] and accuracy = [0.9978, 0.9579] improvements of 2x and 14x, respectively, compared with other methods. The ICP classifier enhanced by artifact detection showed 0.9252 average accuracy with the ground-truth sets provided for validation. Discussion: Results indicate that our approach ensures sturdiness to experimental protocol maneuvers, besides it is effective, simple, and configurable for different studies that use unidimensional time dependent signals as data. Furthermore, our approach would also be effective to analyze the ICPs generated by other cell types, other dyes, chemical stimulation or even signals recorded at higher frequency.
| S-EPMC10213911 | biostudies-literature
Project description:BackgroundMicrobial ecologists now routinely utilize next-generation sequencing methods to assess microbial diversity in the environment. One tool heavily utilized by many groups is the Naïve Bayesian Classifier developed by the Ribosomal Database Project (RDP-NBC). However, the consistency and confidence of classifications provided by the RDP-NBC is dependent on the training set utilized.ResultsWe explored the stability of classification of honey bee gut microbiota sequences by the RDP-NBC utilizing three publically available ribosomal RNA sequence databases as training sets: ARB-SILVA, Greengenes and RDP. We found that the inclusion of previously published, high-quality, full-length sequences from 16S rRNA clone libraries improved the precision in classification of novel bee-associated sequences. Specifically, by including bee-specific 16S rRNA gene sequences a larger fraction of sequences were classified at a higher confidence by the RDP-NBC (based on bootstrap scores).ConclusionsResults from the analysis of these bee-associated sequences have ramifications for other environments represented by few sequences in the public databases or few bacterial isolates. We conclude that for the exploration of relatively novel habitats, the inclusion of high-quality, full-length 16S rRNA gene sequences allows for a more confident taxonomic classification.
| S-EPMC3520854 | biostudies-literature
Project description:It is now widely recognized that the isolation of circulating tumor cells based on cell surface markers might be hindered by variability in their protein expression. Especially in pancreatic cancer, isolation based only on EpCAM expression has produced very diverse results. Methods that are independent of surface markers and therefore independent of phenotypical changes in the circulating cells might increase CTC recovery also in pancreatic cancer. We compared an EpCAM-dependent (IsoFlux) and a size-dependent (automated Siemens Healthineers filtration device) isolation method for the enrichment of pancreatic cancer CTCs. The recovery rate of the filtration based approach is dramatically superior to the EpCAM-dependent approach especially for cells with low EpCAM-expression (filtration: 52%, EpCAM-dependent: 1%). As storage and shipment of clinical samples is important for centralized analyses, we also evaluated the use of frozen diagnostic leukapheresis (DLA) as source for isolating CTCs and subsequent genetic analysis such as KRAS mutation detection analysis. Using frozen DLA samples of pancreatic cancer patients we detected CTCs in 42% of the samples by automated filtration.
| S-EPMC5689673 | biostudies-literature